The A of B

tahnan came up with the silly idea of finding all his iTunes song titles of the form "The A of B", and then sorting the As and Bs separately. So what the heck, I did it with my mp3s:

( 60 made-up song titles and 2 real ones )

Capitalization and punctuation provide some clues for reconstructing the orignal titles, but probably most of them are too obscure to be recognizable (even I don't recognize some of them).

This is how I did it, by the way (in tcsh on Linux):

find . -name "* - The * of *.mp3" -print | sed 's/^.*- //' | sed 's/.mp3//' | grep '^The' > the-of.txt
sed 's/ of .*//' the-of.txt | sort >! the.txt
sed 's/The .* of/of/' the-of.txt | sort >! of.txt
paste -d' ' the.txt of.txt >! the-of-sorted.txt

Perl experts could probably do it in one line, but I'm too old-school.

Update: Oops, I forgot, here's the list of artists: ( 57 artists with some repeats )